Information processing method, device, system, equipment, medium and program product

Abstract

The invention discloses an information processing method, device, system and equipment, a medium and a program product, and is applied to the technical field of artificial intelligence. The method is applied to a cloud server side and comprises the steps that an image recognition request sent by a detection side is received, and the image recognition request is used for requesting the cloud server side to recognize whether a to-be-recognized image comprises contraband goods or not; image features of the to-be-recognized image are recognized through at least one first type of intelligent agent in the multiple first type of intelligent agents, a reference result corresponding to the at least one first type of intelligent agent is obtained, and the reference result comprises a result that whether the to-be-recognized image comprises contraband goods or not; generating a decision result based on the reference result through the second type of agents; and sending the decision result to the detection end. Therefore, the accuracy of the contraband judgment result can be improved.

Description

Technical Field

[0001] The present application belongs to the field of artificial intelligence technology, and in particular, relates to an information processing method, device, system, equipment, medium and program product. Background Art

[0002] In security inspections at public places such as airports and train stations, security inspectors are required to manually review security inspection images to determine whether there are prohibited items in the security inspection images. However, the manual determination of whether there are prohibited items in security inspection images is affected by the differences in the experience levels of security inspectors, which reduces the accuracy of the prohibited items determination results. Summary of the invention

[0003] The embodiments of the present application provide an information processing method, apparatus, system, device, medium and program product, which can improve the accuracy of contraband judgment results.

[0004] In a first aspect, an embodiment of the present application provides an information processing method, which is applied to a cloud service end. The information processing method includes:

[0005] Receive an image recognition request sent by the detection end, the image recognition request carries the image to be recognized, and the image recognition request is used to request the cloud service end to recognize whether the image to be recognized includes contraband;

[0006] Recognize, by at least one of the plurality of first-type intelligent agents, image features of the image to be recognized, and obtain a reference result corresponding to the at least one first-type intelligent agent, the reference result including a result of whether the image to be recognized includes contraband;

[0007] Generate a decision result based on the reference result by the second type of intelligent agent, wherein the decision result includes a result for characterizing whether the image to be identified includes contraband;

[0008] Send the decision result to the detection end.

[0009] In a second aspect, an embodiment of the present application provides an information processing method, which is applied to a detection end, where the detection end includes a detection device and a data gateway. The information processing method includes:

[0010] Through the data gateway, based on the image to be identified collected by the detection device, an image recognition request is sent to the cloud server, and the image recognition request is used to request the cloud server to identify whether the image to be identified includes contraband;

[0011] Receive a decision result sent by the cloud server, the decision result including a result for characterizing whether the image to be identified includes contraband, the decision result is that the cloud server identifies the image features of the image to be identified by at least one first-class intelligent agent among multiple first-class intelligent agents, obtains a reference result corresponding to at least one first-class intelligent agent, and generates a reference result based on the reference result by a second-class intelligent agent, the reference result including a result of whether the image to be identified includes contraband;

[0012] Display the decision results.

[0013] In a third aspect, an embodiment of the present application provides an information processing device, which is applied to a cloud service end. The information processing device includes:

[0014] A first receiving module is used to receive an image recognition request sent by the detection end, the image recognition request carries the image to be recognized, and the image recognition request is used to request the cloud service end to recognize whether the image to be recognized includes contraband;

[0015] A first recognition module is used to recognize image features of an image to be recognized by at least one first-class intelligent agent among a plurality of first-class intelligent agents, and obtain a reference result corresponding to at least one first-class intelligent agent, wherein the reference result includes a result of whether the image to be recognized includes contraband;

[0016] A first generating module is used to generate a decision result based on the reference result by the second type of intelligent agent, wherein the decision result includes a result for characterizing whether the image to be identified includes contraband;

[0017] The first sending module is used to send the decision result to the detection end.

[0018] In a fourth aspect, an embodiment of the present application provides an information processing device, which is applied to a detection end, the detection end includes a detection device and a data gateway, and the information processing device includes:

[0019] A second sending module is used to send an image recognition request to the cloud server through the data gateway based on the image to be recognized collected by the detection device, and the image recognition request is used to request the cloud server to recognize whether the image to be recognized includes contraband;

[0020] A second receiving module is used to receive a decision result sent by the cloud server, the decision result including a result used to characterize whether the image to be identified includes contraband, the decision result is that the cloud server identifies the image features of the image to be identified by at least one first-class intelligent agent among multiple first-class intelligent agents, obtains a reference result corresponding to at least one first-class intelligent agent, and generates a reference result based on the reference result by the second-class intelligent agent, the reference result including a result of whether the image to be identified includes contraband;

[0021] The display module is used to display the decision results.

[0022] In a fifth aspect, an embodiment of the present application provides an information processing system, the information processing system comprising:

[0023] A cloud service end, for implementing the information processing method as described in any one of the first aspects;

[0024] The detection end is connected to the cloud service communication, and is used to implement the information processing method as described in any one of the second aspects.

[0025] In a sixth aspect, an embodiment of the present application provides an electronic device, comprising: a processor and a memory storing computer program instructions; when the processor executes the computer program instructions, it implements an information processing method as described in any one of the first aspects; or it executes an information processing method as described in any one of the second aspects.

[0026] In the seventh aspect, an embodiment of the present application provides a computer-readable storage medium, on which computer program instructions are stored. When the computer program instructions are executed by a processor, an information processing method such as any one of the first aspect is implemented; or an information processing method such as any one of the second aspect is executed.

[0027] In an eighth aspect, an embodiment of the present application provides a computer program product, which includes a computer program or instructions, and when the computer program or instructions are executed by a processor, an information processing method such as any one of the first aspect is implemented; or an information processing method such as any one of the second aspect is executed.

[0028] The information processing method, device, system, equipment, medium and program product of the embodiment of the present application, through at least one first-class intelligent agent among multiple first-class intelligent agents to perform image feature recognition on the image to be identified, the reference result obtained can more accurately distinguish the image feature differences in the image to be identified, and improve the accuracy of the contraband judgment result. Then, based on the reference result obtained by the at least one first-class intelligent agent, a decision result is generated, and further comprehensive processing of the reference result is achieved. Compared with the method of manually judging whether there are contraband in the security inspection image, the decision result obtained by using the first-class intelligent agent and the second-class intelligent agent to comprehensively judge the image to be identified will not be affected by the experience level, and the accuracy of the contraband judgment result can be improved. BRIEF DESCRIPTION OF THE DRAWINGS

[0029] In order to more clearly illustrate the technical solution of the embodiments of the present application, the following is a brief introduction to the drawings required for use in the embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without any creative work.

[0030] Figure 1 A schematic diagram showing a flow chart of an information processing method provided by some embodiments of the present application applied to a cloud service end;

[0031] Figure 2 A flowchart showing a specific implementation of step 130 provided in some embodiments of the present application is shown;

[0032] Figure 3 A flowchart showing a specific implementation of step 1302 provided in some embodiments of the present application is shown;

[0033] Figure 4 A flowchart showing a specific implementation of adjusting a reference weight in a reference weight set in an information processing method provided in some embodiments of the present application is shown;

[0034] Figure 5 A schematic flow chart showing a method for determining a decision result in an information processing method provided in some embodiments of the present application;

[0035] Figure 6 A schematic diagram showing a flow chart of an information processing method provided by some embodiments of the present application applied to a detection terminal;

[0036] Figure 7 A flow chart showing a method for sending a weight adjustment request to a cloud service end in the information processing method provided in some embodiments of the present application;

[0037] Figure 8 A schematic diagram showing a structure of an information processing device provided in some embodiments of the present application applied to a cloud service end;

[0038] Fig. 9 A schematic diagram showing a structure of an information processing device provided in some embodiments of the present application applied to a detection end;

[0039] Fig.10 A schematic diagram of a system for information processing provided by some embodiments of the present application is shown;

[0040] Fig.11 A schematic diagram showing a flow chart of an information processing method provided by some embodiments of the present application applied to an information processing system;

[0041] Fig.12 A schematic diagram of the structure of an electronic device provided in some embodiments of the present application is shown. DETAILED DESCRIPTION

[0042] The features and exemplary embodiments of various aspects of the present application will be described in detail below. In order to make the purpose, technical solutions and advantages of the present application clearer, the present application will be further described in detail below in conjunction with the accompanying drawings and specific embodiments. It should be understood that the specific embodiments described herein are only intended to explain the present application, rather than to limit the present application. For those skilled in the art, the present application can be implemented without the need for some of these specific details. The following description of the embodiments is only to provide a better understanding of the present application by illustrating the examples of the present application.

[0043] It should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, the elements defined by the statement "include..." do not exclude the existence of other identical elements in the process, method, article or device including the elements.

[0044] It should be noted that the acquisition, storage, use and processing of data in the embodiments of the present application are in compliance with the relevant provisions of national laws and regulations.

[0045] It should be noted that in the embodiments of the present application, certain software, components, models and other existing solutions in the industry may be mentioned, and they should be regarded as exemplary. Their purpose is only to illustrate the feasibility of implementing the technical solution of the present application, but it does not mean that the applicant has or will necessarily use the solution.

[0046] Before explaining the technical solutions provided by the embodiments of the present application, in order to facilitate the understanding of the embodiments of the present application, the present application first specifically describes the related technologies involved.

[0047] In security inspections in public places such as airports and train stations, security inspectors are relied upon to manually review security inspection images to determine whether there are prohibited items in the security inspection images. However, the experience levels of security inspectors vary, resulting in different security inspectors' ability to judge prohibited items. For example, some inexperienced security inspectors may find it difficult to accurately identify disguised prohibited items, while experienced security inspectors may be able to identify disguised prohibited items with their rich experience. Therefore, the aforementioned manual review of security inspection images to determine whether there are prohibited items in the security inspection images will result in a prohibited item judgment result that is affected by the differences in the experience levels of security inspectors, resulting in a low accuracy in the prohibited item identification result.

[0048] In order to solve the above-mentioned problems in the related art, the embodiments of the present application provide a method, apparatus, system, device, medium and program product for information processing. Figure 1 To Attachment Figure 7 , the information processing method provided by the embodiment of the present application is described in detail through specific embodiments and their application scenarios.

[0049] Figure 1 The following is a flow chart showing the application of the information processing method provided by some embodiments of the present application to a cloud service end. Figure 1 As shown, the information processing method is applied to a cloud service end, and the information processing method may include steps 110 to 140.

[0050] Step 110: receiving an image recognition request sent by the detection end, the image recognition request carries the image to be recognized, and the image recognition request is used to request the cloud service end to recognize whether the image to be recognized includes contraband.

[0051] Step 120, using at least one first-category intelligent agent among multiple first-category intelligent agents, identifies image features of the image to be identified, and obtains a reference result corresponding to at least one first-category intelligent agent, wherein the reference result includes a result of whether the image to be identified includes contraband.

[0052] Step 130, generating a decision result based on the reference result by the second type of intelligent agent, wherein the decision result includes a result for characterizing whether the image to be identified includes contraband.

[0053] Step 140, sending the decision result to the detection end.

[0054] Thus, by performing image feature recognition on the image to be identified by at least one of the multiple first-type intelligent agents, the reference result obtained can more accurately distinguish the image feature differences in the image to be identified, thereby improving the accuracy of the contraband judgment result. Then, based on the reference result obtained by the at least one first-type intelligent agent, a decision result is generated, thereby achieving further comprehensive processing of the reference result. Compared with the method of manually judging whether there are contraband in the security inspection image, the decision result obtained by using the first-type intelligent agent and the second-type intelligent agent to comprehensively judge the image to be identified will not be affected by the experience level, thereby improving the accuracy of the contraband judgment result.

[0055] The above steps are described in detail below, as shown below.

[0056] First, S110 is involved. The detection end involved in the embodiment of the present application includes a detection device and a data gateway that is communicatively connected to the detection device. The detection device may include, but is not limited to, image acquisition devices such as security inspection machines and cameras. For example, the detection device may be installed in places where security inspections are required, such as airports, railway stations, subway stations, ports, large event venues, entrances to important facilities, etc., to collect images. The function of the detection device is to collect images, i.e., images to be identified. Images to be identified are images that require image feature recognition and analysis, and may include various objects, scenes or information, such as images to be identified may include security inspection images of personnel, luggage, parcels, goods, etc. The data gateway is communicatively connected to the detection device via a 5G data link, so that after the detection device collects the images to be identified, the collected images can be sent to the data gateway via the 5G data link for storage and processing. For example, the security inspection machine collects the images to be identified and sends the images to be identified to the data gateway via the 5G data link. Contraband refers to items that are prohibited from being carried, transported or used, such as explosive, flammable, corrosive, toxic, strongly acidic and alkaline, and radioactive volatile dangerous goods.

[0057] The image recognition request involved in the embodiment of the present application is generated by the detection end based on the image to be recognized, and is used to request the cloud service end to recognize the image to be recognized and determine whether it contains contraband.

[0058] Secondly, with respect to step 120, in the embodiment of the present application, the selection of at least one first-class agent for recognizing the image to be recognized can be determined based on a user-preset rule, such as the recognition task of the image to be recognized can be performed according to the first-class agent associated with the detection end. In this way, it is ensured that the recognition process can be flexibly configured according to the task requirements of the detection end, thereby improving the pertinence and accuracy of the recognition.

[0059] In some embodiments of the present application, at least two of the multiple first-category intelligent agents have different algorithms for recognizing image features, which helps to extract and analyze the image features of the image to be identified from multiple angles, thereby increasing the comprehensiveness and accuracy of the identification of contraband in the image to be identified.

[0060] Exemplarily, each of the multiple agents includes at least one algorithm model for identifying an image to be identified, and the algorithm model can be constructed based on a target detection algorithm. For example, the n agents can include at least two of the following: an agent based on a region with CNN features (R-CNN) series model with convolutional neural network features, an agent based on a YOLO series model, an agent based on a DETR series model, an agent based on a single shot multi-box detector (SSD) model, an agent based on a RetinaNet model, an agent based on an EfficientDet model, an agent based on a Transformer detector (DETR) model, an agent based on a CornerNet model, and an agent based on a CenterNet model.

[0061] Next, the above-mentioned types of intelligent agents will be further described in detail.

[0062] Agents based on the R-CNN family of models. These agents use models such as R-CNN, Fast Regions with CNN features (Fast R-CNN), and Faster Regions with CNN features (Faster R-CNN) to accurately detect and classify contraband in images through region proposal methods and convolutional neural network feature extraction.

[0063] An agent based on the YOLO series of models. This type of agent treats target detection as a regression problem and directly predicts bounding boxes and category probabilities through a single forward propagation, with extremely high detection speed and accuracy.

[0064] Agents based on the SSD model, which combines the speed of YOLO and the accuracy of Faster R-CNN by using filters of different sizes on feature maps of different resolutions to predict bounding boxes and class probabilities. As a result, this type of agent has good detection accuracy while maintaining a high detection speed.

[0065] Agents based on the RetinaNet model, where RetinaNet uses Focal loss to solve the extreme foreground-background class imbalance problem and detects contraband of different scales through a feature pyramid network. As a result, this type of agent can accurately detect contraband of various scales in complex scenes.

[0066] Agents based on the EfficientDet model, where EfficientDet is a multi-scale object detection algorithm that improves detection accuracy and efficiency through compound scaling and progressive refinement. As a result, this type of agent has low computational complexity while maintaining high detection accuracy.

[0067] Agents based on the DETR model, where DETR is a Transformer-based detector that converts the object detection task into a collective prediction problem, eliminating duplicate bounding box predictions. As a result, this type of agent performs well in object detection tasks and can accurately detect and locate objects in images.

[0068] An agent based on the CornerNet model, where CornerNet and CenterNet are object detection algorithms based on corner points and center points. CornerNet uses two parallel networks to predict the corner points and center points of the object respectively.

[0069] An agent based on the CenterNet model, where CenterNet is also an object detection algorithm based on corner points and center points. CenterNet achieves object detection by predicting the center point, width and height of the object.

[0070] It is worth noting here that multiple intelligent agents may include the various models mentioned above, and may also include other algorithms and models to adapt to different detection needs and scenarios.

[0071] In the embodiment of the present application, at least one first-class agent is at least one specific instance of the above-mentioned multiple agents, and the image feature refers to a specific attribute or pattern that can be identified and analyzed in the image, such as color, shape, texture, etc. The reference result includes a reference result corresponding to each first-class agent in the at least one first-class agent, and the reference result corresponding to each first-class agent is a specific result obtained after the corresponding first-class agent identifies the image feature, including information about whether the image to be identified includes contraband.

[0072] In some embodiments of the present application, taking the multiple first agents as n first agents as an example, the image features of the image to be identified can be identified by the i-th first-class agent among the n first-class agents to obtain a reference result, the reference result including the i-th reference result corresponding to the i-th first-class agent, at least two of the n first-class agents have different algorithms for identifying image features, the i-th reference result includes the result of identifying whether the image to be identified includes contraband under the algorithm for identifying image features of the i-th first-class agent, i∈[1, 2, ..., n], wherein n is a positive integer greater than 1, and the n first-class agents refer to n first-class agents participating in the image feature identification task of the image to be identified. i is an integer from 1 to n, used to identify different agents. The n first-class agents perform image feature identification and analysis using different algorithms. In this way, performing image recognition on the same image to be identified by using at least two algorithms for identifying image features among the n first-class agents helps to extract and analyze the image features of the image to be identified from multiple angles, thereby increasing the comprehensiveness and accuracy of identifying contraband in the image to be identified. Then, based on the reference results obtained by these n first-class intelligent agents, a decision result is generated, which realizes a comprehensive judgment of the n reference results. This can reduce the possible recognition errors of the reference results obtained by a single first-class intelligent agent and improve the accuracy of identifying contraband.

[0073] Furthermore, with respect to step 130, the second type of agent refers to an agent that processes the reference results obtained by the preliminary identification of the first type of agent to generate a final decision, such as a training-push integrated server. Exemplarily, the second type of agent may be an agent based on an integrated model, in which the output of each first type of agent is assigned a weight, and the sum of these weights is used to determine the final decision result. Thus, a comprehensive analysis of the reference results output by multiple first type of agents is achieved.

[0074] In some embodiments of the present application, a reference weight set is stored in the second-class agent, and the reference weight set includes a reference weight corresponding to each first-class agent in a plurality of first-class agents. The reference weight corresponding to each first-class agent represents the degree of influence of the output result of each first-class agent on the decision result output by the second-class agent. Based on this, Figure 2 As shown, the above step 130 may specifically include step 1301 to step 1302.

[0075] Step 1301, through the second type of agent, based on the first type of target agent corresponding to the reference result, select the reference weight corresponding to the first type of target agent from the reference weight set.

[0076] The first type of target agent refers to the first type of agent that is associated with the reference result or generates the reference result.

[0077] Exemplarily, according to the first type of target agent, a reference weight corresponding to the first type of target agent is found from the reference weight set.

[0078] Step 1302, generating a decision result based on the reference result and the reference weight corresponding to the first type of target intelligent agent.

[0079] Exemplarily, the decision result is calculated using the screened reference weights and reference results.

[0080] Therefore, by comprehensively considering the reference results output by multiple first-class intelligent agents and the reference weights corresponding to the first-class target intelligent agents corresponding to the reference results, the differences in the recognition results of different first-class intelligent agents for the same image to be recognized can be effectively balanced to obtain a more accurate decision result.

[0081] In some embodiments of the present application, the reference result includes at least one of the following: a first reference result and a second reference result, wherein the first reference result includes a result for characterizing that the image to be identified includes contraband, and the second reference result includes a result for characterizing that the image to be identified does not include contraband. Based on this, the above step 1302 may specifically include steps 13021 to 13023.

[0082] Step 13021, when the reference results include at least one first reference result and at least one second reference result, the reference weights of the first type of target agents corresponding to the first reference results are summed to obtain a first weight value of the first reference result.

[0083] Step 13022, summing up the reference weights of the first type of target agents corresponding to the second reference result to obtain a second weight value of the second reference result.

[0084] Step 13023: determine the reference result corresponding to the target weight value as the decision result, and the target weight value is the largest weight value between the first weight value and the second weight value.

[0085] Next, the above steps 13021 to 13023 will be described in detail with reference to specific examples.

[0086] For example, the number of first-class agents is 5, namely a, b, c, d, e, and the reference weight assigned to each first-class agent by the second-class agent is 20%. When the detection device with device ID set001 sends an image to be identified to the cloud service end through the data gateway, the first-class agents a, b, c, d, e respectively identify the contraband in the image to be identified, among which the first-class agents a, b, c identify the contraband in the image to be identified, and the first-class agents d and e do not identify the contraband in the image to be identified. The first-class agents a, b, c, d, e send their respective reference results to the second-class agents. In the reference results obtained, the second-class agents sum the weights corresponding to the first-class target agents that identify the contraband, that is, a+b+c=20%+20%+20%; and sum the weights corresponding to the first-class target agents that do not identify the contraband, that is, d+e=20%+20%. Thus, the first reference result, i.e., the first weight value of the contraband is 60%, and the second reference result, i.e., the second weight value of the contraband is 40%, is obtained. Based on this, the first reference result corresponding to 60% is determined as the decision result, that is, the decision result is the result that the image to be identified includes contraband.

[0087] Therefore, the reference weight reflects the relative importance of the first type of agent in the process of identifying contraband in the image to be identified, which directly affects the decision result, that is, the reference result with a large weight value has a greater impact on the decision result. Therefore, taking the largest weight value as the decision result can more accurately reflect whether the image to be identified includes contraband, thereby improving the accuracy of the contraband judgment result.

[0088] In some embodiments of the present application, the first reference result also includes a first category reference result and a second category reference result, the first category reference result includes a first type of prohibited goods, the second category reference result includes a second type of prohibited goods, the first weight value includes a first sub-weight value and a second sub-weight value, the first sub-weight value includes a weight value corresponding to the reference weight of the first category target intelligent entity corresponding to the first category reference result, and the second sub-weight value includes a weight value corresponding to the reference weight of the first category target intelligent entity corresponding to the second category reference result.

[0089] Taking the case where the number of first-category agents is 5, namely a, b, c, d, and e, and the reference weight assigned by the second-category agents to each first-category agent is 20% as an example, when the detection device with device ID set001 sends an image to be identified to the cloud server through the data gateway, the first-category agents a, b, c, d, and e respectively identify contraband on the image to be identified. If the reference result obtained by the first-category agent a indicates that contraband exists in the image to be identified, and the type of contraband is controlled knives; the reference result obtained by the first-category agent b indicates that contraband exists in the image to be identified, and the type of contraband is firearms and ammunition; the reference result obtained by the first-category agent c indicates that contraband exists in the image to be identified, and the type of contraband is controlled knives; the reference result obtained by the first-category agent d indicates that contraband exists in the image to be identified, and the type of contraband is controlled knives; the reference result obtained by the first-category agent e indicates that there are no contraband in the image to be identified. In this way, the first sub-weight value is a+c+d=20%+20%+20%, the second sub-weight value is b=20%, and the second weight value is e=20%. Thus, the first-category reference result, i.e., the first sub-weight value of the prohibited item identified as a controlled knife, is 60%, the second-category reference result, i.e., the second sub-weight value of the prohibited item identified as a firearm and ammunition, is 20%, and the second-category reference result, i.e., the second weight value of the prohibited item not identified, is 20%. Based on this, the first-category reference result corresponding to 60% is determined as the decision result, i.e., the decision result is the result that the image to be identified includes prohibited items, and the type of the prohibited item is a controlled knife.

[0090] Therefore, by classifying contraband into different types such as the first contraband type and the second contraband type, and determining the decision result based on the sub-weight value corresponding to each contraband type, not only the result of whether the image to be identified includes contraband, but also the type information of the contraband type is included, thereby improving the accuracy and comprehensiveness of contraband identification.

[0091] In some embodiments of the present application, when the reference results include only one first reference result, the first reference result is determined as the decision result.

[0092] In other embodiments of the present application, when the reference results include at least one first-category reference result and at least one second-category reference result, the reference weights of the first-category target intelligent entities corresponding to the first-category reference results are summed to obtain a third weight value of the first-category reference results, and the reference weights of the first-category target intelligent entities corresponding to the second-category reference results are summed to obtain a fourth weight value of the second-category reference results; then, the reference result corresponding to the largest weight value between the third weight value and the fourth weight value is determined as the decision result.

[0093] In some further embodiments of the present application, when the first reference result includes at least one second reference result, the second reference result is determined as the decision result.

[0094] Then, in step 140, the detection end receives and analyzes the decision result, and then performs corresponding operations, such as alarming, marking contraband, etc., to prompt the detection end to adopt corresponding response strategies based on the decision result.

[0095] In some embodiments of the present application, after step 140, as Figure 4 As shown, the information processing method also includes steps 210 to 220.

[0096] Step 210: receiving feedback information sent by the detection end.

[0097] Among them, the feedback information includes information about whether the decision result is accurate, which is sent by the detection end to the cloud service end after the detection end executes the corresponding response strategy based on the decision result.

[0098] Step 220, when the feedback information includes negative feedback information corresponding to the decision result, adjust at least one of the following: lower the reference weight of the first type of intelligent agent corresponding to the decision result in the reference weight set, and increase the reference weight of the first type of intelligent agent corresponding to the reference result different from the decision result in the reference result.

[0099] Exemplarily, lowering the reference weight of the first type of intelligent agent that generates negative feedback information can reduce the influence of the first type of intelligent agent that generates negative feedback information on the decision-making of the second type of intelligent agent; while increasing the reference weight of the first type of intelligent agent that does not generate negative feedback information can increase the influence of the first type of intelligent agent that does not generate negative feedback information on the decision-making of the second type of intelligent agent, thereby effectively improving the accuracy of the decision results generated by the second type of intelligent agent.

[0100] For example, the number of first-class agents is 5, namely a, b, c, d, and e, and the reference weight assigned to each first-class agent by the second-class agent is 20%. The reference results obtained by the n first-class agents include that the first-class agents a, b, and c have identified contraband in the image to be identified, and the first-class agents d and e have not identified contraband in the image to be identified. In this way, the decision result obtained by the second-class agent is that the contraband is identified in the image to be identified. When receiving the negative feedback information based on the decision result sent by the detection end, that is, the information that there is no contraband in the image to be identified, the reference weights of the first-class agents a, b, and c are reduced from 20% by 1% to 19%, and the reference weights of the first-class agents d and e are increased by 20% by 1% to 21%, so as to improve the accuracy of the decision result obtained by the second-class agent after receiving the new reference result.

[0101] Therefore, when the feedback information includes negative feedback information corresponding to the decision result, that is, the decision result obtained by the second type of intelligent agent is inaccurate, adjusting the reference weight of the first type of intelligent agent corresponding to the decision result in the first type of intelligent agent can optimize the decision process of the second type of intelligent agent and reduce the possibility of similar negative feedback in the subsequent decision-making process of the second type of intelligent agent.

[0102] In some embodiments of the present application, the decision result includes a result and a contraband parameter for characterizing that the image to be identified includes contraband, and the contraband parameter includes at least one of the following: the type of contraband, the position of the contraband in the image to be identified. Based on this, after step 210, the information processing method further includes, when the feedback information includes negative feedback information corresponding to the contraband parameter, adjusting at least one of the following: lowering the reference weight of the first type of intelligent agent corresponding to the contraband parameter in the reference weight set, and raising the reference weight of the first type of intelligent agent corresponding to the reference result that does not include the contraband parameter in the reference result.

[0103] The reference results that do not include the prohibited item parameter include the second reference result and the first reference result that does not include the prohibited item parameter.

[0104] Exemplarily, for the first-class agents whose reference results include incorrect prohibited item parameters, their corresponding reference weights will be reduced. For example, if the decision result incorrectly indicates that the image to be identified includes a controlled knife, and the feedback information indicates that there is no such type of contraband in the image to be identified, then the reference weights of all first-class agents that obtain a controlled knife as the identification result will be lowered. At the same time, for the first-class agents whose reference results do not include incorrect contraband parameters, i.e., prohibited item parameters that do not conform to the negative feedback information, their corresponding reference weights will be increased.

[0105] Referring to the above example, the reference results obtained by the five first-class agents include that the first-class agent a recognizes that the image to be recognized includes contraband, and the type of the contraband is a controlled knife, the first-class agent b recognizes that the image to be recognized includes contraband, and the type of the contraband is a firearm and ammunition, the first-class agent c recognizes that the image to be recognized includes contraband, and the type of the contraband is a controlled knife, the first-class agent d recognizes that the image to be recognized does not include contraband, and the first-class agent e recognizes that the image to be recognized includes contraband, and the type of the contraband is a controlled knife. Based on this, the decision result is that the image to be recognized includes contraband, and the type of the contraband is a controlled knife. However, the feedback information received subsequently indicates that there is actually no controlled knife in the image to be recognized. Therefore, according to the above adjustment rules, the reference weights of the first-class agents a, c, and e can be lowered because a, c, and e mistakenly recognize the controlled knife, and the reference weights of the first-class agents d and b can be increased.

[0106] Therefore, by introducing the feedback adjustment mechanism, the second type of intelligent agent can more accurately consider the actual performance of each first type of intelligent agent when making decisions after receiving new reference results, thereby improving the accuracy of the decision results.

[0107] In some embodiments of the present application, after step 210, the information processing method further includes, when the feedback information includes negative feedback information corresponding to the contraband parameter, and the contraband parameter includes the type of contraband and the position of the contraband in the image to be identified, adjusting at least one of the following:

[0108] The reference weight of the first type of intelligent agent corresponding to the contraband parameter in the reference weight set is adjusted down by a first preset value, and the reference weight of the first type of intelligent agent corresponding to the reference result that does not include the contraband parameter in the reference result is adjusted up by a second preset value;

[0109] Alternatively, when the feedback information includes negative feedback information corresponding to the prohibited item parameter, and the prohibited item parameter includes any one of the type of prohibited item and the position of the prohibited item in the image to be identified, at least one of the following items is adjusted:

[0110] The reference weight of the first type of intelligent agent corresponding to the contraband parameter in the reference weight set is adjusted down by a third preset value, and the reference weight of the first type of intelligent agent corresponding to the reference result that does not include the contraband parameter in the reference result is adjusted up by a fourth preset value;

[0111] Among them, the first preset value is greater than the third preset value, and the second preset value is greater than the fourth preset value.

[0112] Exemplarily, the first preset value is greater than the third preset value, and the second preset value is greater than the fourth preset value, to effectively ensure that when the feedback information is more specific, i.e., includes the type and location of the contraband, the reference weight is adjusted more significantly.

[0113] Referring to the above example, the first type of agents a, b, c, d, e respectively identified the prohibited items on the image to be identified and obtained the type and location of the prohibited items. If the reference results obtained by the first type of agents a, b, c, d, e include:

[0114] a. The image to be identified contains contraband, the type of the contraband is a controlled knife, and the location of the contraband is in the upper left corner of the image to be identified;

[0115] b. The image to be identified includes contraband, the type of contraband is firearms and ammunition, and the location of the contraband is in the lower right corner of the image to be identified;

[0116] c. The image to be identified contains contraband, the type of the contraband is a controlled knife, and the location of the contraband is in the upper left corner of the image to be identified;

[0117] d. The image to be identified contains contraband, the type of the contraband is a controlled knife, and the location of the contraband is at the center of the image to be identified;

[0118] e. The image to be identified does not include contraband.

[0119] If the reference weight assigned by the second-class agent to each first-class agent is 20%, then the weight of identifying that the image to be identified includes contraband, the type of contraband is a controlled knife, and the location of the contraband is in the upper left corner of the image to be identified is a+c=20%+20%=40%; the weight of identifying that the image to be identified includes contraband, the type of contraband is firearms and ammunition, and the location of the contraband is in the lower right corner of the image to be identified is b=20%; the weight of identifying that the image to be identified does not include contraband is e=20%. Based on the weight results, the decision result is that the image to be identified includes contraband, the type of contraband is a controlled knife, and the location of the contraband is in the upper left corner of the image to be identified.

[0120] If the feedback information includes negative feedback information for the recognition result of the controlled tool in the upper left corner of the image to be recognized, adjusting the reference weight set based on the above steps includes: lowering the reference weights of a and c by a first preset value, for example, from 20% to 15% by 5%; and increasing the reference weight of e by a second preset value, for example, from 20% to 25% by 5%. If the feedback information only includes negative feedback information for the recognition result of the controlled tool, adjusting the reference weight set based on the above steps includes: lowering the reference weights of a, c, and d by a third preset value, for example, from 20% to 18% by 2%; and increasing the reference weight of e by a fourth preset value, for example, from 20% to 22%.

[0121] Therefore, when the feedback information includes the type and location of the contraband, it means that the feedback information is more accurate and specific. Therefore, by increasing the adjustment range of the reference weight, that is, the first preset value is greater than the third preset value, and the second preset value is greater than the fourth preset value, this accuracy can be more effectively reflected, thereby helping the second type of intelligent agent to obtain a more accurate decision result.

[0122] In some embodiments of the present application, the image recognition request also carries an application scenario identifier of the detection end. The application scenario identifier is used to identify the application scenario or purpose identifier to which the image recognition request belongs. Based on this, before the above step 140, if Figure 5 As shown, the information processing method may further include steps 310 to 330.

[0123] Step 310: Acquire a target task type associated with the scene identifier according to the association information between the preset scene identifier and the preset task type.

[0124] The association information between the preset scenario identifier and the preset task type is pre-set information that associates different application scenario identifiers with different task types. The target task type includes the target type that needs to be identified according to the application scenario identifier, such as the type of contraband.

[0125] Exemplarily, according to the preset association information, the target task type matching the application scenario identifier in the current image recognition request is found to determine what the specific target task type that the current image recognition request needs to identify is.

[0126] Step 320, adjust the reference result according to the target task type to obtain an adjusted reference result, the adjusted reference result includes an adjusted reference result corresponding to each first-class intelligent agent in at least one first-class intelligent agent, and the adjusted reference result corresponding to each first-class intelligent agent includes the result of identifying whether the image to be identified includes contraband matching the target task type under the algorithm for identifying image features of each first-class intelligent agent. Exemplarily, the initial reference result may include all possible types of contraband, which is a relatively broad and comprehensive list of contraband. After determining the target task type, the initial reference result can be adjusted in a targeted manner. Taking subway security and airplane security as examples, some items may be legal in the subway, but are considered contraband on an airplane. For example, a small volume of drinking water is usually allowed to be carried in the subway, but on an airplane, for safety reasons, liquids exceeding a certain volume are usually not allowed to be carried. If the application scenario identifier points to aircraft security, then after obtaining the reference result, first screen out the prohibited item types that directly match the aircraft security check. For the prohibited item types that are not related to the aircraft security check, adjust the corresponding reference result to the result that the image to be identified does not include the prohibited item, so as to obtain a relatively concise and targeted reference result.

[0127] Step 330, determining the adjusted reference result as the reference result.

[0128] Therefore, by introducing the application scenario identifier and determining the target task type based on the association information between the application scenario identifier and the task type, the adjusted reference result obtained by adjusting the reference result based on the target task type is a more accurate reference result that is closely related to the current task type, which effectively ensures that the reference results obtained by the first type of intelligent agent can achieve the best effect in different application scenarios, thereby improving the accuracy and efficiency of contraband identification.

[0129] It is worth noting here that in the embodiments of the present application, the description of each embodiment has its own focus. For the parts that are not described in detail in a certain embodiment, please refer to the detailed description of the information processing method in the context, and will not be repeated here.

[0130] In some embodiments of the present application, the information processing method applies a detection end, which includes a detection device and a data gateway. Figure 6 As shown, the information processing method may include steps 410 to 430.

[0131] Step 410, through the data gateway, based on the image to be identified collected by the detection device, an image recognition request is sent to the cloud server, and the image recognition request is used to request the cloud server to identify whether the image to be identified includes contraband.

[0132] The detection equipment includes equipment for collecting images to be identified, such as cameras, scanners, etc. The data gateway connects the detection equipment with the cloud service end for data transmission and reception. The cloud service end includes a remote server that provides computing resources and contraband identification services.

[0133] Step 420, receiving the decision result sent by the cloud server, the decision result includes a result for characterizing whether the image to be identified includes contraband, and the decision result is that the cloud server identifies the image features of the image to be identified through at least one first-class intelligent agent among multiple first-class intelligent agents, obtains a reference result corresponding to at least one first-class intelligent agent, and generates a reference result based on the reference result through a second-class intelligent agent, and the reference result includes a result of whether the image to be identified includes contraband.

[0134] Step 430, display the decision result.

[0135] Exemplarily, the detection end also includes electronic devices, including but not limited to computers, computers, tablet computers, and smart phones. Based on this, the decision results can not only be displayed directly on the security inspection equipment, but also can be displayed through electronic devices associated with the security inspection equipment. Through these diverse terminals, the decision results can be conveniently received and viewed. Whether it is the staff at the security inspection site or the remote monitoring and management personnel, they can choose the most suitable terminal device to receive and view the decision information according to their actual needs and permissions.

[0136] Thus, the detection end collects the image to be identified and transmits it to the cloud service end through the data gateway, uses at least one first-type intelligent agent and one second intelligent agent for collaborative identification, and finally generates a decision result, making full use of the computing resources and contraband identification capabilities of the cloud service end, and improving the accuracy and efficiency of identifying contraband in the image to be identified. Finally, the decision result is presented to the user, so that the user can intuitively understand whether the image to be identified contains contraband, which is convenient for subsequent processing and decision-making.

[0137] In some embodiments of the present application, the decision result includes a result that the image to be identified includes contraband and contraband parameters, where the contraband parameters include at least one of the following: the type of contraband, the location of the contraband in the image to be identified; or, the decision result includes a result that the image to be identified does not include contraband.

[0138] As a result, users can quickly and intuitively understand whether the image to be identified includes contraband, as well as the type of contraband and the location of the contraband in the image to be identified, so that they can respond more quickly, effectively shortening the time it takes for users to perform tasks and improving task execution efficiency.

[0139] In some embodiments of the present application, Figure 7 As shown, the information processing method may further include steps 510 to 530.

[0140] Step 510: upon receiving the user's input on the decision result, obtaining feedback information corresponding to the decision result.

[0141] Among them, input refers to the information provided by the user's actions when interacting with the detection terminal, including but not limited to clicking a "satisfied" or "unsatisfied" button, entering specific feedback information, and uploading images or videos related to the image to be identified and contraband.

[0142] Step 520, when the feedback information includes negative feedback information corresponding to the decision result, a weight adjustment request is generated based on the negative feedback information, and the weight adjustment request is used to request the cloud server to adjust the reference weights in the reference weight set stored in the second type of intelligent agent based on the negative feedback information.

[0143] Step 530: Send a weight adjustment request to the cloud service end.

[0144] Therefore, by obtaining user feedback information and sending a weight adjustment request to the cloud server based on negative feedback information, the cloud server can adjust the reference weights in the reference weight set stored by the second agent to optimize the decision-making effect of the second agent.

[0145] Based on the information processing method provided in the above embodiment, the present application also provides a specific implementation of the information processing device. Please refer to the following embodiment.

[0146] See also Figure 8 The information processing device 600 provided in the embodiment of the present application is applied to a cloud service end, and the information processing device 600 includes:

[0147] The first receiving module 610 is used to receive an image recognition request sent by the detection end, the image recognition request carries the image to be recognized, and the image recognition request is used to request the cloud service end to recognize whether the image to be recognized includes contraband;

[0148] The first recognition module 620 is used to recognize the image features of the image to be recognized by at least one first-type intelligent agent among the plurality of first-type intelligent agents, and obtain a reference result corresponding to the at least one first-type intelligent agent, wherein the reference result includes a result of whether the image to be recognized includes contraband;

[0149] A first generating module 630 is used to generate a decision result based on the reference result by the second type of intelligent agent, wherein the decision result includes a result for indicating whether the image to be identified includes contraband;

[0150] The first sending module 640 is used to send the decision result to the detection end.

[0151] Thus, the image recognition request carrying the image to be recognized is received by the first receiving module 610, and the image recognition is performed on the same image to be recognized by using at least one first-class intelligent agent in the first recognition module 620. The reference result obtained can more accurately distinguish the image feature differences in the image to be recognized, thereby improving the accuracy of the judgment result of contraband. Then, the first generating module 630 is used to generate a decision result based on the reference result obtained by the at least one first-class intelligent agent, thereby achieving further comprehensive processing of the reference result. Finally, the first sending module 640 is used to send the decision result to the detection end to prompt the user of the detection end whether the image to be recognized includes contraband. Compared with the method of manually judging whether there are contraband in the security inspection image, the decision result obtained by using the first-class intelligent agent and the second-class intelligent agent to comprehensively judge the contraband will not be affected by the experience level, and can improve the accuracy of the judgment result of contraband.

[0152] In some embodiments of the present application, the first generating module 630 may specifically include:

[0153] A first screening submodule is used for, when a reference weight set is stored in the second-class agent and the reference weight set includes a reference weight corresponding to each of the first-class agents in the plurality of first-class agents, screening, by the second-class agent, a reference weight corresponding to the first-class target agent from the reference weight set based on the first-class target agent corresponding to the reference result;

[0154] The first generating submodule is used to generate a decision result according to the reference result and the reference weight corresponding to the first type of target intelligent agent.

[0155] In some embodiments of the present application, the first generation submodule may specifically include:

[0156] A first processing unit is configured to sum the reference weights of the first type of target agents corresponding to the first reference results to obtain a first weight value of the first reference result when the reference results include at least one first reference result and at least one second reference result, the first reference result includes a result for representing that the image to be identified includes contraband, and the second reference result includes a result for representing that the image to be identified does not include contraband;

[0157] A second processing unit, configured to sum the reference weights of the first type of target agents corresponding to the second reference result to obtain a second weight value of the second reference result;

[0158] The first determining unit is used to determine a reference result corresponding to a target weight value as a decision result, where the target weight value is the largest weight value between the first weight value and the second weight value.

[0159] In some embodiments of the present application, when the first reference result in the first processing unit also includes a first category reference result and a second category reference result, the first category reference result includes a first type of prohibited goods, and the second category reference result includes a second type of prohibited goods, the first weight value includes a first sub-weight value and a second sub-weight value, and the first sub-weight value includes a weight value corresponding to the reference weight of the first category target intelligent entity corresponding to the first category reference result, and the second sub-weight value includes a weight value corresponding to the reference weight of the first category target intelligent entity corresponding to the second category reference result.

[0160] In some embodiments of the present application, the above-mentioned information processing device may further include:

[0161] A third receiving module, used to receive feedback information sent by the detection end;

[0162] The first adjustment module is configured to adjust at least one of the following items when the feedback information includes negative feedback information corresponding to the decision result:

[0163] The reference weight of the first type of intelligent agent corresponding to the decision result in the reference weight set is lowered, and the reference weight of the first type of intelligent agent corresponding to the reference result different from the decision result in the reference result is increased.

[0164] In some embodiments of the present application, the above-mentioned information processing device may further include:

[0165] The second adjustment module is configured to adjust at least one of the following items when the decision result includes a result for characterizing that the image to be identified includes contraband and a contraband parameter, the contraband parameter includes at least one of the following: a type of contraband, a position of the contraband in the image to be identified, and the feedback information includes negative feedback information corresponding to the contraband parameter:

[0166] The reference weight of the first type of intelligent agent corresponding to the prohibited item parameter in the reference weight set is lowered, and the reference weight of the first type of intelligent agent corresponding to the reference result that does not include the prohibited item parameter in the reference result is increased.

[0167] In some embodiments of the present application, the second adjustment module may also be used to adjust at least one of the following items when the feedback information includes negative feedback information corresponding to the prohibited item parameter, and the prohibited item parameter includes the type of the prohibited item and the position of the prohibited item in the image to be identified:

[0168] The reference weight of the first type of intelligent agent corresponding to the contraband parameter in the reference weight set is adjusted down by a first preset value, and the reference weight of the first type of intelligent agent corresponding to the reference result that does not include the contraband parameter in the reference result is adjusted up by a second preset value;

[0169] Alternatively, when the contraband parameter includes any one of the parameters of the contraband type and the position of the contraband in the image to be identified, at least one of the following items is adjusted:

[0170] The reference weight of the first type of intelligent agent corresponding to the contraband parameter in the reference weight set is adjusted down by a third preset value, and the reference weight of the first type of intelligent agent corresponding to the reference result that does not include the contraband parameter in the reference result is adjusted up by a fourth preset value;

[0171] Among them, the first preset value is greater than the third preset value, and the second preset value is greater than the fourth preset value.

[0172] In some embodiments of the present application, the above-mentioned information processing device may further include:

[0173] A first acquisition module is used to acquire a target task type associated with the scene identifier according to association information between a preset scene identifier and a preset task type before generating a decision result step based on a reference result by a second type of intelligent agent when the image recognition request also carries an application scene identifier of the detection end;

[0174] A third adjustment module is used to adjust the reference result according to the target task type to obtain an adjusted reference result, wherein the adjusted reference result includes an adjusted reference result corresponding to each first-class intelligent agent in at least one first-class intelligent agent, and the adjusted reference result corresponding to each first-class intelligent agent includes a result of identifying whether the image to be identified includes contraband matching the target task type under the algorithm for identifying image features of each first-class intelligent agent;

[0175] The first determining module is used to determine the adjusted reference result as the reference result.

[0176] Each module of the information processing device 600 provided in the embodiment of the present application can realize Figures 1 to 5The functions of each step of the provided information processing method and its ability to achieve corresponding technical effects are described briefly and will not be elaborated here.

[0177] See also Fig. 9 The information processing device 700 provided in the embodiment of the present application is applied to a detection end, and the detection end includes a detection device and a data gateway. The information processing device 700 may include:

[0178] The second sending module 710 is used to send an image recognition request to the cloud server through the data gateway based on the image to be recognized collected by the detection device, and the image recognition request is used to request the cloud server to recognize whether the image to be recognized includes contraband;

[0179] The second receiving module 720 is used to receive the decision result sent by the cloud server, the decision result includes a result for characterizing whether the image to be identified includes contraband, and the decision result is that the cloud server identifies the image features of the image to be identified by at least one first-class intelligent agent among multiple first-class intelligent agents, obtains a reference result corresponding to at least one first-class intelligent agent, and generates a reference result based on the reference result by the second-class intelligent agent, the reference result includes a result of whether the image to be identified includes contraband;

[0180] The display module 730 is used to display the decision result.

[0181] Thus, the second sending module 710 is used to send the image to be identified based on the detection end through the data gateway to the cloud service end, so that the cloud service end uses at least one first-type intelligent agent and a second intelligent agent for collaborative identification, and finally generates a decision result, which fully utilizes the computing resources and contraband identification capabilities of the cloud service end, and improves the accuracy and efficiency of identifying contraband in the image to be identified. Finally, the decision result is received by the second receiving module 720, and the decision result is presented to the user through the display module 730, so that the user can intuitively understand whether the image to be identified contains contraband, which is convenient for subsequent processing and decision-making.

[0182] In some embodiments of the present application, the decision result in the above-mentioned display module 730 includes the result that the image to be identified includes contraband and contraband parameters, and the contraband parameters include at least one of the following: the type of contraband, the position of the contraband in the image to be identified; or, the decision result includes the result that the image to be identified does not include contraband.

[0183] In some embodiments of the present application, the information processing device may further include:

[0184] A second acquisition module is used to acquire feedback information corresponding to the decision result when receiving the user's input on the decision result;

[0185] A second generating module is used to generate a weight adjustment request based on the negative feedback information when the feedback information includes negative feedback information corresponding to the decision result, and the weight adjustment request is used to request the cloud server to adjust the reference weight in the reference weight set stored in the second type of intelligent agent based on the negative feedback information;

[0186] The third sending module is used to send a weight adjustment request to the cloud server.

[0187] Each module of the information processing device 700 provided in the embodiment of the present application can realize Figure 6 and Figure 7 The functions of each step of the provided information processing method and its ability to achieve corresponding technical effects are described briefly and will not be elaborated here.

[0188] The embodiment of the present application provides a schematic diagram of an information processing system. The information processing system includes a cloud service end and a detection end, wherein the cloud service end is used to implement Figures 1 to 5 A detection terminal, connected to a cloud service for communication, is used to implement Figure 6 to Figure 7 information processing methods.

[0189] In some embodiments of the present application, Fig.10 As shown, the detection end may include a detection device 1010, a data gateway 1020 and an electronic device 1030. Exemplarily, the detection device 1010 may include a plurality of security inspection devices of different models, which are deployed at a plurality of security inspection locations. The data gateway is used to manage and transmit the security inspection images generated by these security inspection devices, i.e., the images to be identified. The data gateway 1020 may include a plurality of detection device side intelligent terminals, which are used to receive security inspection images from each security inspection device.

[0190] In some embodiments of the present application, Fig.10 As shown, the cloud service end may include n first-type agents 1040 and second-type agents 1050. In order to ensure the efficiency and stability of data transmission, the security inspection images and application scenario identifiers related to the security inspection images received by the data gateway may be sent to the n first-type agents for processing using a 5G data link or dedicated network to obtain a reference result, and then the reference result is sent to the second-type agent so that the second-type agent generates a decision result based on the reference result. Then, the decision result is displayed on the electronic device.

[0191] In some embodiments of the present application, the present application also provides an information processing method, the execution subject of the information processing method is Fig.10 The information processing system shown is Fig.11 As shown, the information processing method may include steps 801 to 813.

[0192] Step 801: The detection device collects an image to be identified.

[0193] Step 802: The detection device sends the image to be identified to the data gateway.

[0194] Among them, when the detection device sends the image to be identified, it carries the first task information, and the first task information includes the task identifier, the device identifier, the task type, the location information of the detection device, the data type and the timestamp sequence. Specifically, the task identifier is an ID that uniquely identifies a task and is used to distinguish different tasks. The device identifier is the unique ID of the detection device, which is used to identify which detection device sent the image. The task type is used to describe the nature of the task, such as image recognition, contraband identification, etc. The location information of the detection device refers to the location data of the detection device, such as longitude and latitude. The data type refers to the data type of the image to be identified. The timestamp sequence refers to the time when the image is acquired, which is used for sorting or analyzing time correlation.

[0195] It can be understood that through task identification and device identification, the execution of tasks can be tracked and managed; location information and timestamp sequences can be used for subsequent data analysis, such as analyzing image recognition requirements in different time periods or locations; task types and data types help the cloud server determine how to process or assign tasks.

[0196] Step 803: The data gateway generates an image recognition request based on the image to be recognized.

[0197] Step 804: The data gateway sends an image recognition request to n first-category agents.

[0198] When sending the image recognition request, the data gateway carries the second task information, which includes the first task information and the data gateway identifier. Specifically, the data gateway identifier is a unique ID of the data gateway, which is used to identify which data gateway sent the image recognition request.

[0199] Step 805: n first-category intelligent agents identify whether the image to be identified includes contraband and obtain a reference result.

[0200] Step 806: n first-category agents send reference results to second-category agents.

[0201] When the n first-class agents send reference results to the second-class agents, they also carry the second task information and the identifier of each of the n first-class agents. Specifically, the identifier of each of the n first-class agents is used by the second-class agent to identify which first-class agent sent the specific reference result.

[0202] Step 807: The second type of agent generates a decision result based on the reference result and the reference weight of the first type of target agent corresponding to the reference result. The decision result includes a result for characterizing whether the image to be identified includes contraband.

[0203] Step 808: The second type of agent sends the decision result to the electronic device.

[0204] When the second-type agent sends the decision result to the electronic device, it also carries the task identifier, the second-type agent identifier, the reference result, and the timestamp sequence. Specifically, the second-type agent identifier is used by the electronic device to identify which second-type agent sent the decision result, so as to track or verify the output of the second-type agent.

[0205] Step 809: The electronic device receives and displays the decision result.

[0206] Step 810: When receiving the user's input on the decision result, the electronic device obtains feedback information corresponding to the decision result.

[0207] Step 811: When the feedback information includes negative feedback information corresponding to the decision result, the electronic device generates a weight adjustment request based on the negative feedback information.

[0208] Step 812: The electronic device sends a weight adjustment request to the second type of agent.

[0209] The weight adjustment request carries feedback information, task identification, electronic device identification, and timestamp sequence. Specifically, the electronic device identification is used by the second type of agent to identify which electronic device sent the weight adjustment request to ensure the accuracy and security of communication.

[0210] Step 813: The second type of agent adjusts the reference weight in the reference weight set stored in the second type of agent based on the negative feedback information.

[0211] Thus, through the parallel processing of multiple first-class agents and the comprehensive decision-making of the second-class agents, the efficiency and accuracy of identifying contraband in the image to be identified are improved. In addition, through user feedback and weight adjustment mechanism, the information processing system can be continuously optimized to further improve the accuracy of contraband identification.

[0212] Fig.12 A schematic diagram of the hardware structure of an electronic device provided in some embodiments of the present application is shown.

[0213] The electronic device 900 may include a processor 901 and a memory 902 storing computer program instructions.

[0214] Specifically, the processor 901 may include a central processing unit (CPU), or an application specific integrated circuit (ASIC), or may be configured to implement one or more integrated circuits of the embodiments of the present application.

[0215] The memory 902 may include a large capacity memory for data or instructions. By way of example and not limitation, the memory 902 may include a hard disk drive (HDD), a floppy disk drive, a flash memory, an optical disk, a magneto-optical disk, a magnetic tape, or a universal serial bus (USB) drive or a combination of two or more of these. In appropriate cases, the memory 902 may include a removable or non-removable (or fixed) medium. In appropriate cases, the memory 902 may be inside or outside the integrated gateway disaster recovery device. In a specific embodiment, the memory 902 is a non-volatile solid-state memory.

[0216] In a particular embodiment, the memory 902 may include a read-only memory (ROM), a random access memory (RAM), a magnetic disk storage medium device, an optical storage medium device, a flash memory device, an electrical, optical or other physical / tangible memory storage device. Therefore, generally, the memory 902 includes one or more tangible (non-transitory) computer-readable storage media (e.g., a memory device) encoded with software including computer-executable instructions, and when the software is executed (e.g., by one or more processors), it is operable to perform the operations described with reference to the data processing method according to the first aspect of the present application.

[0217] The processor 901 implements any one of the information processing methods in the above embodiments by reading and executing computer program instructions stored in the memory 902 .

[0218] In one example, the electronic device may further include a communication interface 903 and a bus 910. Fig. 9 As shown, the processor 901, the memory 902, and the communication interface 903 are connected via a bus 910 and communicate with each other.

[0219] The communication interface 903 is mainly used to implement communication between various modules, devices, units and / or equipment in the embodiments of the present application.

[0220] Bus 910 includes hardware, software or both, and the parts of electronic equipment are coupled to each other.For example, but not limitation, bus may include accelerated graphics port (AGP) or other graphics bus, enhanced industrial standard architecture (EISA) bus, front side bus (FSB), hypertransport (HT) interconnection, industrial standard architecture (ISA) bus, infinite bandwidth interconnection, low pin count (LPC) bus, memory bus, micro channel architecture (MCA) bus, peripheral component interconnection (PCI) bus, PCI-Express (PCI-X) bus, serial advanced technology attachment (SATA) bus, video electronics standard association local (VLB) bus or other suitable bus or two or more of these combinations. In appropriate cases, bus 910 may include one or more buses. Although the present application embodiment describes and shows a specific bus, the application considers any suitable bus or interconnection.

[0221] The electronic device can execute the information processing method in the embodiment of the present application, thereby realizing the combination Figures 1 to 11 A method and apparatus for information processing are described.

[0222] In addition, in combination with the information processing method in the above-mentioned embodiment, the embodiment of the present application may provide a computer-readable storage medium for implementation. The computer-readable storage medium stores computer program instructions; when the computer program instructions are executed by the processor, any one of the information processing methods in the above-mentioned embodiments is implemented. Examples of computer-readable storage media include non-transitory computer-readable storage media, such as portable disks, hard disks, random access memories (RAM), read-only memories (ROM), erasable programmable read-only memories (EPROM or flash memory), portable compact disk read-only memories (CD-ROM), optical storage devices, magnetic storage devices, etc.

[0223] In addition, in combination with the information processing method in the above embodiment, the embodiment of the present application may provide a computer program product for implementation. The program product is stored in a storage medium, and may specifically include a computer program or instruction, which implements any one of the information processing methods in the above embodiment when executed by a processor. The program product is executed by at least one processor to implement the various processes of the above data processing method embodiment, and can achieve the same technical effect, so it will not be repeated here to avoid repetition.

[0224] It should be clear that the present application is not limited to the specific configuration and processing described above and shown in the figures. For the sake of simplicity, a detailed description of the known method is omitted here. In the above embodiments, several specific steps are described and shown as examples. However, the method process of the present application is not limited to the specific steps described and shown, and those skilled in the art can make various changes, modifications and additions, or change the order between the steps after understanding the spirit of the present application.

[0225] The functional blocks shown in the above block diagram can be implemented as hardware, software, firmware or a combination thereof. When implemented in hardware, it can be, for example, an electronic circuit, an application specific integrated circuit (ASIC), appropriate firmware, a plug-in, a function card, etc. When implemented in software, the elements of the present application are programs or code segments that are used to perform the required tasks. The program or code segment can be stored in a machine-readable medium, or transmitted on a transmission medium or a communication link by a data signal carried in a carrier wave. "Machine-readable medium" can include any medium capable of storing or transmitting information. Examples of machine-readable media include electronic circuits, semiconductor memory devices, ROM, flash memory, erasable ROM (EROM), floppy disks, CD-ROMs, optical disks, hard disks, optical fiber media, radio frequency (RF) links, etc. The code segment can be downloaded via a computer network such as the Internet, an intranet, etc.

[0226] It should also be noted that the exemplary embodiments mentioned in this application describe some methods or systems based on a series of steps or devices. However, this application is not limited to the order of the above steps, that is, the steps can be performed in the order mentioned in the embodiment, or in a different order from the embodiment, or several steps can be performed simultaneously.

[0227] Aspects of the present disclosure are described above with reference to the flowchart and / or block diagram of the method, device (system) and computer program product according to the embodiment of the present disclosure. It should be understood that each box in the flowchart and / or block diagram and the combination of each box in the flowchart and / or block diagram can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing device to produce a machine so that these instructions executed by the processor of the computer or other programmable data processing device enable the implementation of the function / action specified in one or more boxes of the flowchart and / or block diagram. Such a processor can be, but is not limited to, a general-purpose processor, a special-purpose processor, a special application processor, or a field programmable logic circuit. It can also be understood that each box in the block diagram and / or flowchart and the combination of boxes in the block diagram and / or flowchart can also be implemented by dedicated hardware that performs a specified function or action, or can be implemented by a combination of dedicated hardware and computer instructions.

[0228] The above is only a specific implementation of the present application. Those skilled in the art can clearly understand that for the convenience and simplicity of description, the specific working processes of the systems, modules and units described above can refer to the corresponding processes in the aforementioned method embodiments, and will not be repeated here. It should be understood that the protection scope of the present application is not limited to this. Any technician familiar with the technical field can easily think of various equivalent modifications or replacements within the technical scope disclosed in this application, and these modifications or replacements should be included in the protection scope of this application.

Claims

1. A method of information processing, characterized in that: Applied to a cloud server, the method includes: Receive an image recognition request sent by the detection end, the image recognition request carries an image to be recognized, and the image recognition request is used to request the cloud service end to recognize whether the image to be recognized includes contraband; Recognize the image features of the image to be recognized by at least one first-class intelligent agent among the plurality of first-class intelligent agents, and obtain a reference result corresponding to the at least one first-class intelligent agent, wherein the reference result includes a result of whether the image to be recognized includes contraband; Generate a decision result based on the reference result by a second type of intelligent agent, wherein the decision result includes a result for characterizing whether the image to be identified includes contraband; Send the decision result to the detection end.

2. The method according to claim 1, characterized in that The second-type agent stores a reference weight set, wherein the reference weight set includes a reference weight corresponding to each first-type agent in the plurality of first-type agents; The generating of a decision result based on the reference result by the second type of intelligent agent includes: By means of the second type of agents, based on the first type of target agents corresponding to the reference results, selecting reference weights corresponding to the first type of target agents from the reference weight set; A decision result is generated according to the reference result and the reference weight corresponding to the first type of target intelligent agent.

3. The method according to claim 2, characterized in that The reference result includes a reference result corresponding to each first-category intelligent agent in the at least one first-category intelligent agent, and the reference result includes at least one of the following: a first reference result and a second reference result, wherein the first reference result includes a result for characterizing that the image to be identified includes contraband, and the second reference result includes a result for characterizing that the image to be identified does not include the contraband; The step of generating a decision result according to the reference result and the reference weight corresponding to the first type of target intelligent agent includes: In a case where the reference result includes at least one first reference result and at least one second reference result, summing the reference weights of the first type of target agents corresponding to the first reference results to obtain a first weight value of the first reference result; Sum the reference weights of the first type of target agents corresponding to the second reference result to obtain a second weight value of the second reference result; A reference result corresponding to a target weight value is determined as the decision result, wherein the target weight value is the largest weight value between the first weight value and the second weight value.

4. The method according to claim 3, characterized in that The first reference result also includes a first category of reference results and a second category of reference results, the first category of reference results includes a first type of prohibited items, the second category of reference results includes a second type of prohibited items, the first weight value includes a first sub-weight value and a second sub-weight value, the first sub-weight value includes a weight value corresponding to a reference weight of a first category of target intelligent entity corresponding to the first category of reference results, and the second sub-weight value includes a weight value corresponding to a reference weight of a first category of target intelligent entity corresponding to the second category of reference results.

5. The method according to claim 2, characterized in that: The method further comprises: Receiving feedback information sent by the detection end; In the case where the feedback information includes negative feedback information corresponding to the decision result, adjusting at least one of the following: The reference weight of the first type of intelligent agent corresponding to the decision result in the reference weight set is lowered, and the reference weight of the first type of intelligent agent corresponding to the reference result different from the decision result in the reference result is increased.

6. The method according to claim 5, characterized in that The decision result includes a result for characterizing that the image to be identified includes contraband and contraband parameters, wherein the contraband parameters include at least one of the following: the type of contraband, the position of the contraband in the image to be identified; After receiving the feedback information sent by the detection end, the method further includes: In the case where the feedback information includes negative feedback information corresponding to the contraband parameter, adjusting at least one of the following: The reference weight of the first type of intelligent agent corresponding to the prohibited item parameter in the reference weight set is lowered, and the reference weight of the first type of intelligent agent corresponding to the reference result that does not include the prohibited item parameter in the reference result is increased.

7. The method according to claim 6, characterized in that After receiving the feedback information sent by the detection end, the method further includes: In a case where the feedback information includes negative feedback information corresponding to the contraband parameter, and the contraband parameter includes the type of contraband and the position of the contraband in the image to be identified, adjusting at least one of the following: The reference weight of the first type of intelligent agent corresponding to the contraband parameter in the reference weight set is adjusted down by a first preset value, and the reference weight of the first type of intelligent agent corresponding to the reference result that does not include the contraband parameter in the reference result is adjusted up by a second preset value; Alternatively, when the feedback information includes negative feedback information corresponding to the contraband parameter, and the contraband parameter includes any one of the type of contraband and the position of the contraband in the image to be identified, at least one of the following items is adjusted: The reference weight of the first type of intelligent agent corresponding to the contraband parameter in the reference weight set is adjusted down by a third preset value, and the reference weight of the first type of intelligent agent corresponding to the reference result that does not include the contraband parameter in the reference result is adjusted up by a fourth preset value; Among them, the first preset value is greater than the third preset value, and the second preset value is greater than the fourth preset value.

8. The method according to claim 1, characterized in that The image recognition request also carries an application scenario identifier of the detection end; Before generating a decision result based on the reference result by the second type of intelligent agent, the method further includes: According to the association information between the preset scene identifier and the preset task type, obtaining the target task type associated with the scene identifier; Adjusting the reference result according to the target task type to obtain an adjusted reference result, wherein the adjusted reference result includes an adjusted reference result corresponding to each first-category intelligent agent in the at least one first-category intelligent agent, and the adjusted reference result corresponding to each first-category intelligent agent includes a result of identifying whether the image to be identified includes contraband matching the target task type under the algorithm of each first-category intelligent agent for identifying the image features; The adjusted reference result is determined as the reference result.

9. The method according to any one of claims 1 to 8, characterized in that: At least two of the plurality of first-category intelligent agents have different algorithms for recognizing the image features.

10. A method of information processing, characterized in that: Applied to a detection end, the detection end includes a detection device and a data gateway, and the method includes: Sending an image recognition request to a cloud server through the data gateway based on the image to be recognized collected by the detection device, wherein the image recognition request is used to request the cloud server to recognize whether the image to be recognized includes contraband; receiving a decision result sent by the cloud server, the decision result including a result for characterizing whether the image to be identified includes contraband, the decision result being that the cloud server identifies image features of the image to be identified by at least one first-class intelligent agent among a plurality of first-class intelligent agents, obtains a reference result corresponding to the at least one first-class intelligent agent, and generates a reference result based on the reference result by a second-class intelligent agent, the reference result including a result of whether the image to be identified includes contraband; The decision result is displayed.

11. The method according to claim 10, characterized in that The decision result includes the result that the image to be identified includes contraband and contraband parameters, and the contraband parameters include at least one of the following: the type of contraband, the position of the contraband in the image to be identified; Alternatively, the decision result includes a result that the image to be identified does not include the contraband.

12. The method according to claim 10 or 11, characterized in that: The method further comprises: Upon receiving a user input on the decision result, obtaining feedback information corresponding to the decision result; In a case where the feedback information includes negative feedback information corresponding to the decision result, generating a weight adjustment request based on the negative feedback information, wherein the weight adjustment request is used to request the cloud service end to adjust a reference weight in a reference weight set stored in the second type of agent based on the negative feedback information; Send the weight adjustment request to the cloud service end.

13. An information processing device, characterized in that: Applied to a cloud server, the device comprises: A first receiving module, configured to receive an image recognition request sent by a detection end, wherein the image recognition request carries an image to be recognized, and the image recognition request is used to request the cloud service end to recognize whether the image to be recognized includes contraband; A first recognition module, configured to recognize image features of the image to be recognized by at least one first-type intelligent agent among a plurality of first-type intelligent agents, and obtain a reference result corresponding to the at least one first-type intelligent agent, wherein the reference result includes a result of whether the image to be recognized includes contraband; A first generating module, configured to generate a decision result based on the reference result by a second type of intelligent agent, wherein the decision result includes a result for characterizing whether the image to be identified includes contraband; The first sending module is used to send the decision result to the detection end.

14. An information processing device, characterized in that: Applied to a detection end, the detection end includes a detection device and a data gateway, and the device includes: A second sending module, configured to send an image recognition request to a cloud server through the data gateway based on the image to be recognized collected by the detection device, wherein the image recognition request is used to request the cloud server to recognize whether the image to be recognized includes contraband; A second receiving module is used to receive a decision result sent by the cloud server, the decision result including a result used to characterize whether the image to be identified includes contraband, the decision result is that the cloud server identifies the image features of the image to be identified by at least one first-class intelligent agent among multiple first-class intelligent agents, obtains a reference result corresponding to the at least one first-class intelligent agent, and generates a reference result based on the reference result by a second-class intelligent agent, the reference result including a result of whether the image to be identified includes contraband; The display module is used to display the decision result.

15. An information processing system, characterized in that: include: A cloud service end, for implementing the information processing method as claimed in any one of claims 1 to 9; The detection end is connected to the cloud service for implementing the information processing method as described in any one of claims 10 to 12.

16. An electronic device, characterized in that: The device comprises: a processor and a memory storing computer program instructions; When the processor executes the computer program instructions, the processor implements the information processing method according to any one of claims 1 to 9; or executes the information processing method according to any one of claims 10 to 12.

17. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer program instructions, which, when executed by a processor, implement the information processing method as described in any one of claims 1 to 9; or execute the information processing method as described in any one of claims 10 to 12.

18. A computer program product, characterized in that The invention comprises a computer program, wherein when the computer program is executed, the computer program realizes the information processing method according to any one of claims 1 to 9; or executes the information processing method according to any one of claims 10 to 12.

Images